Friday, August 25, 2017

Greenspan on interest rates

Here is a recent Bloomberg article, quoting Alan Greenspan. "By any measure, real long-term interest rates are much too low and therefore unsustainable... We are experiencing a bubble, not in stock prices but in bond prices.”

The article goes on more about the Greenspan and the Fed's view, which seems to also be shared by Goldman Sachs Group Inc. Chief Economist David Kostin.  It seems to include the following points:
  • Rising rates hurts stock valuations.
  • High inflation hurts stock valuations.
  • Rising rates are a sign of tight monetary policy.
These three points manage to be both wrong and contradictory.  I suppose one could construct ways that they could all be true.  If inflation shot up to 8%, and the Fed tightened as a reaction to that, then I suppose, inflation would be high, rates would be high, the Fed would be tightening, and equity values would suffer, as in the 1970s.  So, there is some context where these factors would be true.  But, the fact that this is the particular context we are talking about right now, to me, is a sign of our current problem.  Should we really be worried about this?

And, while that hypothetical might salvage the frame of view they are using, it seems clear to me that they do not limit their framework to that specific context.  In the article, even shifts in interest rates and inflation back toward moderate levels are referenced as if they would cause price contractions in equities.

The article does mention that there could be doubts to this view, and that brief peaks in bond rates in 2013 and 2016 didn't cause equities to decline.

This shouldn't be surprising.  This is an example of how an upside-down CAPM model helps to think about these things more clearly.  There is little relationship between expected returns on equities and bond rates.

One reason that it seems people think there is a relationship is because frequently nominal bond rates are compared to implied equity yields.  Equities are a real asset - their values increase over time with inflation.  Most bonds are nominal - their face values are fixed at a nominal value.  I frequently see nominal rates compared to equity returns, and most of the time, this is an incoherent comparison.  To Bloomberg's credit, in the linked article, they refer to TIPS bonds, which are also real assets - adjusted for inflation over time.

But, even using real rates, I just don't see any reason for this view.  Over the nearly 20 year period of time that we have real TIPS rates on long term treasuries (20 and 30 year maturities are what I use here.), there is little relationship between expected equity returns and TIPS rates.

Here, I will use Aswath Damodaran's estimate of the equity risk premium (ERP) to estimate expected returns on equities.  Expected real returns on equities are equal to ERP plus the long term real TIPS rate.  On a quarterly basis back to 2008, after a brief jump in equity yields during the crisis, total real expected equity returns have remained around 6% to 7%.

We can see this even more clearly with annual data, which can take us back to 1999.  Here, again, total real expected returns to equities remain around 6% to 7%.  This is in spite of real rates falling from 4% to less than 1%.

This is the upside down CAPM model.  Start at about 7% real expected returns on equities.  That is stable over time, with some noise.  When TIPS are at 4%, that means investors are willing to give up 3% for safety.  When TIPS are at 2%, investors are willing to give up 5% for safety.  It's as simple as that.  Firms are price-takers when it comes to interest rates.  Interest rates reflect risk aversion.  When they are low, risk aversion is high.  Firms are price-takers in capital markets, so there is no reason to expect them to leverage up and take financial risks when risk aversion is high.  And, in fact, I don't think we see that.  In equilibrium, I'm not sure that leverage is a particularly cyclical factor.

To the extent that very high inflation in the 1970s seemed to increase (decrease) expected equity returns (prices), this may be mostly due to de facto tax rates.  When inflation is high, firms are taxed on booked profits that are really just inflationary.  In other contexts, if PE ratios are changing, we should interpret that as a shift in growth expectations.  Thinking about equities in terms of premiums to treasury yields is a red herring.

9 comments:

  1. Like many bubble predictions, this one has been around for years. At least 15 years. Sad.

    ReplyDelete
    Replies
    1. It's crazy how many things are taken as gospel that have such questionable empirical backing.

      Delete
  2. This comment has been removed by the author.

    ReplyDelete
  3. There is a virus or monetary theosphy out there that interest rates are always and everywhere too low. For two generations we have heard about easy money, and watched as inflation and interest rates trended towards zero.

    A bit surprising that Greenspan should be saying so; he was often criticized for being "too loose" yet his record was pretty good. In general, rates were (historically) lowish and growth good in the Greenspan years.

    Of course, IOER is negative in Japan, and the BoJ holds 10-year JGBs at zero yield. Japan has scorching 0.5% inflation. (This almost gives credence to Fisherians, that if a central banks holds rates low then inflation follows.)

    As discussed often on this blog, interest rates and housing prices are very complicated story, with some academics suggesting current account deficits are the key variable. Property zoning too, and we are talking 40% on CPI. Property zoning is a macroeconomic issue.

    A bubble in bonds is also a reprise of the stance that we believe in markets, we believe in efficient markets, except when we don't. If there is a "bond bubble" that means sophisticated institutional buyers of bonds are being duped, another version of the stance that all one needs to do destroy pinstriped captains of industry and turn them into big-pants bozos is lower interest rates.

    Denver housing prices 60% above the bubble era.

    ReplyDelete
    Replies
    1. Yep.

      On Denver, by the way, it appears that Price/Rent levels are about the same as they were at the peak of the boom. It's all rent inflation.

      Delete
  4. Dean Baker once replied to me (about 2 years ago, I think) that the only reason his paper on the housing bubble (dated August 2002) was early, not wrong, was because of low interest rates.

    ReplyDelete
    Replies
    1. I suppose that's true, to some extent. Low rates did help raise prices, and if the Fed had jacked up rates in 2002 to kill the market, I suppose they would have succeeded.

      I wonder what he thinks of those who blame CDOs or private securitizations for the "bubble", since those markets were negligible in 2002.

      Delete
  5. On Denver, is the short story then that the city has "matured" and so now property owners are acting to constrain new construction?

    ReplyDelete
    Replies
    1. There are some political attempts to curb building, and obviously, there are places like Boulder in Colorado. But, I think Denver is building at a reasonable rate.

      Really, home prices in Dallas look similar over this time. We broke the mortgage market, which has been a conduit for new supply. So, cities that are economically successful right now don't have a fully functional conduit for new supply, regardless of their zoning regime.

      Low tier markets are like Germany or Switzerland, where there is less real housing consumption, there aren't tax benefits for it, and there are many renters who depend on investor building. Then, on top of that market, there is the high tier market where households can get financing, and they still enjoy all the American subsidies to owner-occupied housing. So homes that are still in the owner-occupier market reflect those benefits and prices rise along with rents. That's my best guess as to what's going on. Rent is rising, but it doesn't trigger supply like it used to, but meanwhile the same factors that make prices high are still in effect. If we got rid of the non-price rationing of mortgage credit, rents would go down. If we got rid of the owner-occupier tax advantages, prices would go down. But, for now, we have a regime that is pushing both rents and prices up.

      Delete